Avtomatiziranje kontrole kakovosti in dozimetrije z odprtokodnimi programi : magistrsko delo

Bogović, David (author), Žibert, Janez (mentor)

Uvod: Zanesljivo delovanje slikovno-diagnostične opreme je pogoj za uspešno in kakovostno izvajanje teh storitev znotraj ustanove. Odgovornim osebam za nadzor kakovosti je zato treba zagotoviti enostaven dostop in pregled podatkov izvajanja kontrole kakovosti in prejetih doz ionizirajočega sevanja (dozimetrija). Namen: Namen magistrske naloge je izgradnja odprtokodne programske rešitve avtomatiziranja postopkov kvantitativne obdelave podatkov, ki podajajo parametre kakovosti slikovno-zajemalne opreme in izpostavljenost preiskovancev ionizirajočemu sevanju, shranjevanja rezultatov z možnostjo enostavnega dostopa, povzemanja in grafičnega prikazovanja podatkov ter podajanja teh v standardizirani obliki za notranja ali zunanja telesa, ki izvajajo nadzor nad zagotavljanjem kakovosti. Metode dela: Uporabili smo odprtokodna orodja (knjižnice), s katerimi smo izgradili sistem programov, ki avtomatizira in poveže različna področja, prek katerih zagotovimo standardizacijo na področju kontrole kakovosti skupaj z dozimetrijo. Delovanje sistema smo preverili z retrospektivno študijo, s katero primerjamo rezultate analiziranih slik homogenosti kamer gama našega sistema z orodjem IAEA-NMQC Toolkit in evalviranjem točnosti znakovne prepoznave slik CT-doznih poročil. Primerljivosti sistemov za analizo slik smo preverili s testiranjem povprečij vzorcev. Vzorci so vsebovali vrednosti rezultatov analiziranih slik štirih parametrov (integralna in diferencialna homogenost območij UFOV in CFOV) pri različnih oblikah detektorjev. Mejo za statistično značilnost smo postavili pri 0,05. Rezultati: Ugotovili smo, da posamezni programi delujejo po pričakovanjih ter povezujejo sistem v celoto. Z izgrajenim sistemom smo uspešno zagotovili avtomatiziranje postopkov spremljanja kontrole kakovosti in izpostavljenosti preiskovancev ionizirajočem sevanju. Omogočili smo tudi enostaven dostop do rezultatov, njihovo povzemanje in grafično upodabljanje ter izvažanje in poročanje v standardizirani obliki. Primerjava rezultatov homogenosti ne pokaže statistično značilnih razlik med vzorci (p > 0,05). Korelacija med spremenljivkama v vsakem paru posameznega parametra je močno pozitivna za vse oblike detektorjev (R >= 0,965). Točnost znakovne prepoznave CT-dozimetričnih količin je bila 100-odstotna. Razprava in zaključek: Sistem avtomatizacije lahko učinkovito pripomore pri aktivnostih izvajanja kontrole kakovosti in dozimetrije. Najbolj pomembna dejavnika, ki bi ju izpostavili, sta časovna učinkovitost in standardizacija.

• Language: Slovenian
• Keywords: kontrola kakovosti, dozimetrija, avtomatiziranje, standardizacija, odprtokodni programi
• Work type: Master's thesis/paper
• Organization: ZF - Faculty of Health Sciences
• Year: 2019
• PID: 20.500.12556/RUL-109981
• COBISS.SI-ID: 5678955
• Publication date in RUL: 11.09.2019

Secondary language

• Language: English
• Title: Automating quality control and dosimetry with open source software : master's thesis

Abstract

Introduction: Reliable operation of imaging equipment is the requirement for the high-quality provision of healthcare services in the field of medical imaging. Therefore, it is necessary to provide easy access and data review on quality control activities and patient dosimetry to those responsible for quality assurance. Purpose: The purpose of the master's thesis is to build an open source software solution for automating quantitative data processing that provides quality parameters of imaging equipment and exposure of patients to ionizing radiation, database storage for results, which are easily accessible to the user, data analysis of results with various graphical representations, and reporting in a standardized form for internal and external bodies that supervise quality assurance. Methods: We used open source tools (libraries) to construct and develop a system of programs that automate and connect different areas, through which we ensure standardization in the field of quality control and dosimetry. Operation of the system was assessed with retrospective study with which we compared the results of gamma camera uniformity analysis with IAEA-NMQC Toolkit and evaluated optical character recognition accuracy of CT-dose images. Uniformity was tested by comparing means. The comparison was made for every detector shape for four parameters, usually reported as uniformity results (integral and differential uniformity for UFOV and CFOV). The threshold of p-value was set at 0,05. Results: We could confirm that individual programs work as intended and that they connect and give entirety to the system. The system successfully automates the monitoring of quality control and patient radiation exposure. It also provides easy access and analysis of results with standardized reporting. Comparison of uniformity results did not show any significant statistical differences between samples (p > 0.05). The correlation of variables in each pair was strong positive (R >= 0.965) for every parameter. Discussion and conclusion: Automating data processing can effectively contribute to quality control and dosimetry activities. The most important factors to be highlighted are time efficiency and standardization.

• Keywords: quality control, dosimetry, automating, standardization, open source software